SYSTEM AND METHOD USING SEPARABLE TRANSFER LEARNING BASED ARTIFICIAL NEURAL NETWORK

摘要

A computing system using artificial intelligence based on machine learning is disclosed. The computing system may include at least one processor; a first artificial neural network for pre-training a first phase task based on a first dataset; a second artificial neural network for pre-training a task of a second phase based on a second dataset; and a target artificial neural network for training a target task based on a target dataset, wherein the at least one processor sets the parameters of the first phase layer of the first artificial neural network after pre-training the first artificial neural network to the target artificial neural network. transfer to the first phase layer of the neural network, and after pre-training of the second artificial neural network, transfer the parameters of the second phase layer of the second artificial neural network to the second phase layer of the target artificial neural network, and After transferring the parameters of the first phase layer of the first artificial neural network and the parameters of the second phase layer of the second artificial neural network, the target artificial neural network is controlled to fine-tuning the target task.